CN107578048A - A kind of long sight scene vehicle checking method based on vehicle rough sort - Google Patents
A kind of long sight scene vehicle checking method based on vehicle rough sort Download PDFInfo
- Publication number
- CN107578048A CN107578048A CN201710653791.1A CN201710653791A CN107578048A CN 107578048 A CN107578048 A CN 107578048A CN 201710653791 A CN201710653791 A CN 201710653791A CN 107578048 A CN107578048 A CN 107578048A
- Authority
- CN
- China
- Prior art keywords
- vehicle
- bwl
- rectangle
- msub
- mrow
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Abstract
The invention discloses a kind of long sight scene vehicle checking method based on vehicle rough sort, the present invention first carries out rough sort to long sight scene vehicle and obtains candidate's large car and candidate's compact car, then candidate's vehicle is detected with the vehicle window grader of corresponding vehicle, improves the accuracy and real-time of vehicle detection.
Description
Technical field
The invention belongs to wisdom traffic field, is related to a kind of vehicle checking method, more particularly to examine in long sight scene vehicle
During survey, the method to being detected after vehicle progress vehicle rough sort with different classifications device to corresponding vehicle vehicle.
Background technology
At present, vehicle ownership increases rapidly, more than the growth rate of same period urban road and means of transportation, city road
Road jam situation is extremely serious.Traffic congestion can cause the travel time to increase, vehicle start-stop time increases, energy consumption is substantially increased,
Aggravate the negative consequences such as environmental pollution.Vehicle queue length contributes to the car to crossing as the important parameter in intelligent transportation
Flow is predicted, and carries out counter-measure in advance, is reduced the generation of crossing congestion situation, is significant.
In vehicle checking method, include with the immediate technical scheme of the present invention：Publication No. CN104978567A's
Chinese patent application discloses the vehicle checking method based on scene classification, and this method carries out scene classification to input video, obtained
To simple scenario and complex scene.The simple scenario is modeled using average frame background modeling method, to the complicated field
Scape is modeled using Gaussian Background modeling.The prospect binary map obtained to background modeling is screened, and obtains candidate's vehicle
Region.Then Hog and LBP features are extracted to detect vehicle after obtaining grader as cascade classifier feature, training.Should
Road gate monitoring scene is categorized as simple and complex scene by method, then carries out background using different methods to different scenes
Modeling, the effect of foreground extraction is improved, but this method is not classified to candidate's vehicle, and still vehicle is entered using single grader
Row detection；Publication No. CN104239898A Chinese patent application discloses a kind of quick bayonet vehicle comparison and vehicle cab recognition
Method, this method first carry out prospect vehicle detection and obtain vehicle region, then extract SIFT feature description of vehicle region, look into
Realize that vehicle slightly matches after asking model data storehouse, reuse SIFT and candidate's vehicle image is accurately matched, believed by geometry
Final vehicle comparison result is obtained after breath checking.This method needs to establish model data storehouse in advance, extracts sift features and matching
Vehicle storehouse amount of calculation is larger, is not suitable for real-time vehicle detection, and vehicle checking method proposed by the present invention is under high definition scene
Also real-time detection can be reached；Carry out vehicle detection under long sight scene, front and rear part unavoidably occurs in vehicle during congestion in road
Occlusion issue, this is interfered to vehicle detection.In addition, in order to ensure the precision of vehicle queue length, it is necessary to large car and
Compact car carries out vehicle classification.When using background modeling method extraction vehicle foreground, in order to detect oversize vehicle, detection zone
Domain often sets larger, causes vehicle detection time-consuming longer, has a strong impact on the real-time of video analytic system.
The content of the invention
In view of the above-mentioned problems, the invention provides a kind of long sight scene vehicle checking method based on vehicle rough sort.This
Invention comprises the following steps：
Step 1：Grader is trained, gathers the positive and negative samples of the oversize vehicle and dilly vehicle window in long sight scene, point
The HOG features of each positive negative sample are indescribably taken, the vehicle window for training to obtain large car and compact car using SVM classifier is classified
Device；
Step 2：Pretreatment, demarcate the medium-and-large-sized vehicle detection region of long sight scene manually from traffic surveillance videos, then
Demarcate compact car detection zone manually in oversize vehicle detection region；Oversize vehicle is demarcated respectively and dilly vehicle window is high
The minimum threshold of degree, width and area；
Step 3：Image sequence is read, obtains the oversize vehicle detection region M in present frame；
Step 4：Background modeling is carried out to M, obtains prospect binary map；
Step 5：Prospect binary map in step 4 is screened to obtain candidate's vehicle, driving then is entered to candidate's vehicle
Type rough sort, it is specially：
Step 5.1：Medium filtering and expansive working first are carried out to the prospect binary map in step 4, after being handled before
Scape binary map G；
Step 5.2：Dilly detection zone in interception G obtains prospect binary map GS, finds all connected regions in GS
The minimum enclosed rectangle in domain, form compact car vehicle window boundary rectangle set SWL={ swli| i=1,2,3...n }, n represents external
Rectangle number, it is set to meet formula (1), (2) simultaneously：
swli.W ＞ SCar.W and swli.H ＞ SCar.H (1)
swli.S ＞ SCar.S (2)
In formula, swliRepresent i-th of compact car vehicle window boundary rectangle, swli.H、swli.W、swli.S swl is represented respectivelyi's
Highly, width and area；SCar.H, SCar.W, SCar.S represent dilly vehicle window rectangle minimum constructive height, minimum widith respectively
With the threshold value of minimum area；
Step 5.3：The minimum enclosed rectangle of all connected regions in G is found, forms large car vehicle window boundary rectangle set
BWL={ bwli| i=1,2,3...m }, m represents boundary rectangle number, it is met formula (3), (4) simultaneously：
bwli.W ＞ BCar.W and bwli.H ＞ BCar.H (3)
bwli.S ＞ BCar.S (4)
In formula, bwliRepresent i-th of large car vehicle window boundary rectangle, bwli.H、bwli.W、bwli.S bwl is represented respectivelyi's
Highly, width and area, BCar.H, BCar.W, BCar.S represent oversize vehicle vehicle window boundary rectangle minimum constructive height, width respectively
With the threshold value of area；
Step 5.4：Note tracking vehicle boundary rectangle set TL={ tli| i=1,2,3 ..., p }, wherein p is tracking car
Sum, if bwliMeet formula (5) or formula (6), then judge bwliFor false candidate's oversize vehicle, further being rejected from BWL should
Rectangle；This process is repeated, until all boundary rectangles in traversal BWL；
In formula, tlj∩bwliRepresent rectangle tljAnd bwliIntersecting area,Represent the area of intersecting area；tlj.X、
tlj.W rectangle tl is represented respectivelyjUpper left angle point X-coordinate and width；bwli.X、bwlj.W rectangle bwl is represented respectivelyiThe upper left corner
Point X-coordinate and width；tlj.center.Y the Y-coordinate of rectangular centre point is represented, G.Buttom represents oversize vehicle detection region
Rectangular base Y-coordinate；
Step 6：Vehicle detection, candidate's vehicle of corresponding vehicle is detected using the vehicle window grader trained, had
Body is：
Step 6.1：BWL corresponding boundary rectangle subgraphs in oversize vehicle detection region are intercepted, with oversize vehicle vehicle window
Grader detects to the subgraph of interception, obtains pinpoint oversize vehicle vehicle window boundary rectangle set NTLB={ ntlbi
| i=1,2 ..., r }, wherein r represents the vehicle window number detected, ntlbiRepresent i-th of cart vehicle window boundary rectangle；
Step 6.2：Interception SWL corresponds to boundary rectangle subgraph in dilly detection zone, with dilly vehicle window point
Class device detects to the subgraph of interception, obtains pinpoint dilly vehicle window boundary rectangle set NTLS={ ntlsi|i
=1,2 ..., q }, wherein q represents the vehicle window number detected, ntlsiRepresent i-th of dolly vehicle window boundary rectangle；
Step 6.3：If any rectangle ntlb in NTLBiMeet formula (7), then it is assumed that the rectangle is the large car newly detected
, TL is added into, is otherwise rejected；
In formula, ntlbi∩tljRepresent rectangle ntlbiAnd tljIntersecting area,Represent the area of intersecting area；
Step 6.4：If any rectangle ntls in NTLSiMeet formula (8), then it is assumed that the rectangle is small-sized newly to detect
Vehicle, TL is added into, is otherwise rejected；
In formula, ntlsi∩tljRepresent rectangle ntlsiAnd tljIntersecting area,Represent the area of intersecting area；
Step 7：Judge whether current frame number is less than sequence image numbering maximum, if performing step 3 less than going to, otherwise
Terminate；
Advantages of the present invention and beneficial effect are：It is large-scale that the present invention first obtains candidate to long sight scene vehicle progress rough sort
Car and candidate's compact car, then candidate's vehicle is detected with the vehicle window grader of corresponding vehicle, improves vehicle detection
Accuracy and real-time.
Brief description of the drawings
Fig. 1 is a kind of flow chart of the long sight scene vehicle checking method based on vehicle window.
Fig. 2 is the long sight scene frame of video full figure in the embodiment of the present invention.
Fig. 3 is long sight scene large car vehicle detection region in the embodiment of the present invention.
Fig. 4 is long sight scene compact car vehicle detection region in the embodiment of the present invention.
Fig. 5 is the prospect binary map that background modeling obtains in the embodiment of the present invention.
Fig. 6 is vehicle detection result in the embodiment of the present invention.
Embodiment
The long sight scene vehicle checking method based on vehicle rough sort of the present invention is elaborated with reference to embodiment
Embodiment.In the present embodiment, reference picture 1, to a kind of long sight scene vehicle detection side based on vehicle rough sort
Method is specifically introduced：
Step 1：Grader is trained, gathers the positive and negative samples of the oversize vehicle and dilly vehicle window in long sight scene, point
The HOG features of each positive negative sample are indescribably taken, the vehicle window for training to obtain large car and compact car using SVM classifier is classified
Device；In the present embodiment, in actual monitored video for vehicle in the road video monitoring of multiple different scenes, it is artificial to cut
It is the large-scale of 400*400 or so to take compact car vehicle window picture that 2000 pixels are 200*100 or so and 2000 pixel pixels
Car vehicle window, these positive sample pictures should include complete vehicle window and include background as few as possible.The negative sample picture of vehicle is adopted
Collection process is：The collection non-vehicle window picture similar to positive sample size, 4000 conducts are at least chosen in these pictures and bear sample
This；
Step 2：Pretreatment, demarcate the medium-and-large-sized vehicle detection region of long sight scene manually from traffic surveillance videos, then
Demarcate compact car detection zone manually in oversize vehicle detection region；Oversize vehicle is demarcated respectively and dilly vehicle window is high
The minimum threshold of degree, width and area；In the present embodiment, frame of video full figure is as shown in Fig. 2 the oversize vehicle detection area of interception
Domain is as shown in figure 3, dilly detection zone is as shown in Figure 4；Large car vehicle window minimum constructive height and width threshold value be respectively 150,
150 and 23000, compact car vehicle window minimum constructive height and width threshold value are respectively 32,64 and 2500；
Step 3：Image sequence is read, obtains the oversize vehicle detection region M in present frame；In the present embodiment, image
Sequence number maximum is Count；
Step 4：Background modeling is carried out to M, obtains prospect binary map；
Step 5：Prospect binary map in step 4 is screened to obtain candidate's vehicle, driving then is entered to candidate's vehicle
Type rough sort, it is specially：
Step 5.1：Medium filtering and expansive working first are carried out to the prospect binary map in step 4, after being handled before
Scape binary map G；In the present embodiment, the prospect binary map G that background modeling obtains is as shown in Figure 5；
Step 5.2：Dilly detection zone in interception G obtains prospect binary map GS, finds all connected regions in GS
The minimum enclosed rectangle in domain, form compact car vehicle window boundary rectangle set SWL={ swli| i=1,2,3...n }, n represents external
Rectangle number, it is set to meet formula (1), (2) simultaneously：
swli.W ＞ SCar.W and swli.H ＞ SCar.H (1)
swli.S ＞ SCar.S (2)
In formula, swliRepresent i-th of compact car vehicle window boundary rectangle, swli.H、swli.W、swli.S swl is represented respectivelyi's
Highly, width and area；SCar.H, SCar.W, SCar.S represent dilly vehicle window rectangle minimum constructive height, minimum widith respectively
With the threshold value of minimum area；
Step 5.3：The minimum enclosed rectangle of all connected regions in G is found, forms large car vehicle window boundary rectangle set
BWL={ bwli| i=1,2,3...m }, m represents boundary rectangle number, it is met formula (3), (4) simultaneously：
bwli.W ＞ BCar.W and bwli.H ＞ BCar.H (3)
bwli.S ＞ BCar.S (4)
In formula, bwliRepresent i-th of large car vehicle window boundary rectangle, bwli.H、bwli.W、bwli.S bwl is represented respectivelyi's
Highly, width and area, BCar.H, BCar.W, BCar.S represent oversize vehicle vehicle window boundary rectangle minimum constructive height, width respectively
With the threshold value of area；
Step 5.4：Note tracking vehicle boundary rectangle set TL={ tli| i=1,2,3 ..., p }, wherein p is tracking car
Sum, if bwliMeet formula (5) or formula (6), then judge bwliFor false candidate's oversize vehicle, further being rejected from BWL should
Rectangle；This process is repeated, until all boundary rectangles in traversal BWL；
In formula, tlj∩bwliRepresent rectangle tljAnd bwliIntersecting area,Represent the area of intersecting area；tlj.X、
tlj.W rectangle tl is represented respectivelyjUpper left angle point X-coordinate and width；bwli.X、bwlj.W rectangle bwl is represented respectivelyiThe upper left corner
Point X-coordinate and width；tlj.center.Y the Y-coordinate of rectangular centre point is represented, G.Buttom represents oversize vehicle detection region
Rectangular base Y-coordinate；
Step 6：Vehicle detection, candidate's vehicle of corresponding vehicle is detected using the vehicle window grader trained, had
Body is：
Step 6.1：BWL corresponding boundary rectangle subgraphs in oversize vehicle detection region are intercepted, with oversize vehicle vehicle window
Grader detects to the subgraph of interception, obtains pinpoint oversize vehicle vehicle window boundary rectangle set NTLB={ ntlbi
| i=1,2 ..., r }, wherein r represents the vehicle window number detected, ntlbiRepresent i-th of cart vehicle window boundary rectangle；
Step 6.2：Interception SWL corresponds to boundary rectangle subgraph in dilly detection zone, with dilly vehicle window point
Class device detects to the subgraph of interception, obtains pinpoint dilly vehicle window boundary rectangle set NTLS={ ntlsi|i
=1,2 ..., q }, wherein q represents the vehicle window number detected, ntlsiRepresent i-th of dolly vehicle window boundary rectangle；
Step 6.3：If any rectangle ntlb in NTLBiMeet formula (7), then it is assumed that the rectangle is the large car newly detected
, TL is added into, is otherwise rejected；
In formula, ntlbi∩tljRepresent rectangle ntlbiAnd tljIntersecting area,Represent the area of intersecting area；At this
In embodiment, the result of vehicle detection is as shown in Figure 6；
Step 6.4：If any rectangle ntls in NTLSiMeet formula (8), then it is assumed that the rectangle is small-sized newly to detect
Vehicle, TL is added into, is otherwise rejected；
In formula, ntlsi∩tljRepresent rectangle ntlsiAnd tljIntersecting area,Represent the area of intersecting area；
Step 7：Judge whether current frame number is less than sequence image numbering maximum, if performing step 3 less than going to, otherwise
Terminate；
The content only citing to present inventive concept way of realization described in this specification embodiment, protection of the invention
Scope is not construed as being only limitted to the concrete form that embodiment is stated, protection scope of the present invention is also and in art technology
Personnel are according to the thinkable equivalent technologies mean of present inventive concept institute.
Claims (3)
1. a kind of long sight scene vehicle checking method based on vehicle rough sort, comprises the following steps：
Step 1：Grader is trained, the positive and negative samples of the oversize vehicle and dilly vehicle window in long sight scene is gathered, carries respectively
The HOG features of each positive negative sample are taken, train to obtain the vehicle window grader of large car and compact car using SVM classifier；
Step 2：Pretreatment, the medium-and-large-sized vehicle detection region of long sight scene is demarcated manually from traffic surveillance videos, then big
Compact car detection zone is demarcated in type vehicle detection region manually；Oversize vehicle and dilly vehicle window height, width are demarcated respectively
The minimum threshold of degree and area；
Step 3：Image sequence is read, obtains the oversize vehicle detection region M in present frame；In the present embodiment, image sequence
Number maximum is Count；
Step 4：Background modeling is carried out to M, obtains prospect binary map；
Step 5：Prospect binary map in step 4 is screened to obtain candidate's vehicle, it is thick then to carry out vehicle to candidate's vehicle
Classification；
Step 6：Vehicle detection, candidate's vehicle of corresponding vehicle is detected using the vehicle window grader trained
Step 7：Judge whether current frame number is less than sequence image numbering maximum, if performing step 3 less than going to, otherwise tie
Beam.
2. the long sight scene vehicle checking method based on vehicle rough sort as claimed in claim 1, it is characterised in that：Step 5
Specially：
Step 5.1：Medium filtering and expansive working, the prospect two after being handled first are carried out to the prospect binary map in step 4
Value figure G；In the present embodiment, the prospect binary map G that background modeling obtains is as shown in Figure 5；
Step 5.2：Dilly detection zone in interception G obtains prospect binary map GS, finds all connected regions in GS
Minimum enclosed rectangle, form compact car vehicle window boundary rectangle set SWL={ swli| i=1,2,3...n }, n represents boundary rectangle
Number, it is set to meet formula (1), (2) simultaneously：
swli.W ＞ SCar.W and swli.H ＞ SCar.H (1)
swli.S ＞ SCar.S (2)
In formula, swliRepresent i-th of compact car vehicle window boundary rectangle, swli.H、swli.W、swli.S swl is represented respectivelyiHeight
Degree, width and area；SCar.H, SCar.W, SCar.S represent respectively dilly vehicle window rectangle minimum constructive height, minimum widith and
The threshold value of minimum area；
Step 5.3：The minimum enclosed rectangle of all connected regions in G is found, forms large car vehicle window boundary rectangle set BWL=
{bwli| i=1,2,3...m }, m represents boundary rectangle number, it is met formula (3), (4) simultaneously：
bwli.W ＞ BCar.W and bwli.H ＞ BCar.H (3)
bwli.S ＞ BCar.S (4)
In formula, bwliRepresent i-th of large car vehicle window boundary rectangle, bwli.H、bwli.W、bwli.S bwl is represented respectivelyiHeight
Degree, width and area, BCar.H, BCar.W, BCar.S represent respectively oversize vehicle vehicle window boundary rectangle minimum constructive height, width and
The threshold value of area；
Step 5.4：Note tracking vehicle boundary rectangle set TL={ tli| i=1,2,3 ..., p }, wherein p is total for tracking vehicle
Number, if bwliMeet formula (5) or formula (6), then judge bwliFor false candidate's oversize vehicle, the square is further rejected from BWL
Shape；This process is repeated, until all boundary rectangles in traversal BWL；
<mrow>
<msub>
<mi>S</mi>
<mrow>
<msub>
<mi>tl</mi>
<mi>j</mi>
</msub>
<mo>&cap;</mo>
<msub>
<mi>bwl</mi>
<mi>i</mi>
</msub>
</mrow>
</msub>
<mo>></mo>
<mn>0</mn>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>5</mn>
<mo>)</mo>
</mrow>
</mrow>
In formula, tlj∩bwliRepresent rectangle tljAnd bwliIntersecting area,Represent the area of intersecting area；tlj.X、tlj.W
Rectangle tl is represented respectivelyjUpper left angle point X-coordinate and width；bwli.X、bwlj.W rectangle bwl is represented respectivelyiUpper left angle point X
Coordinate and width；tlj.center.Y the Y-coordinate of rectangular centre point is represented, G.Buttom represents oversize vehicle detection region rectangle
Bottom Y-coordinate.
3. the long sight scene vehicle checking method based on vehicle rough sort as claimed in claim 1, it is characterised in that：Step 6
Specially：
Step 6.1：BWL corresponding boundary rectangle subgraphs in oversize vehicle detection region are intercepted, are classified with oversize vehicle vehicle window
Device detects to the subgraph of interception, obtains pinpoint oversize vehicle vehicle window boundary rectangle set NTLB={ ntlbi| i=
1,2 ..., r }, wherein r represents the vehicle window number detected, ntlbiRepresent i-th of cart vehicle window boundary rectangle；
Step 6.2：Interception SWL corresponds to boundary rectangle subgraph in dilly detection zone, with dilly vehicle window grader
The subgraph of interception is detected, obtains pinpoint dilly vehicle window boundary rectangle set NTLS={ ntlsi| i=1,
2 ..., q }, wherein q represents the vehicle window number detected, ntlsiRepresent i-th of dolly vehicle window boundary rectangle；
Step 6.3：If any rectangle ntlb in NTLBiMeet formula (7), then it is assumed that the rectangle is the oversize vehicle newly detected, will
It adds TL, is otherwise rejected；
<mrow>
<msub>
<mi>S</mi>
<mrow>
<msub>
<mi>ntlb</mi>
<mi>i</mi>
</msub>
<mo>&cap;</mo>
<msub>
<mi>tl</mi>
<mi>j</mi>
</msub>
</mrow>
</msub>
<mo>=</mo>
<mn>0</mn>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>7</mn>
<mo>)</mo>
</mrow>
</mrow>
In formula, ntlbi∩tljRepresent rectangle ntlbiAnd tljIntersecting area,Represent the area of intersecting area；In this implementation
In example, the result of vehicle detection is as shown in Figure 6；
Step 6.4：If any rectangle ntls in NTLSiMeeting formula (8), then it is assumed that the rectangle is the dilly newly detected,
TL is added into, is otherwise rejected；
<mrow>
<msub>
<mi>S</mi>
<mrow>
<msub>
<mi>ntls</mi>
<mi>i</mi>
</msub>
<mo>&cap;</mo>
<msub>
<mi>tl</mi>
<mi>j</mi>
</msub>
</mrow>
</msub>
<mo>=</mo>
<mn>0</mn>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>8</mn>
<mo>)</mo>
</mrow>
</mrow>
In formula, ntlsi∩tljRepresent rectangle ntlsiAnd tljIntersecting area,Represent the area of intersecting area.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710653791.1A CN107578048B (en) | 2017-08-02 | 2017-08-02 | Vehicle type rough classification-based far-view scene vehicle detection method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710653791.1A CN107578048B (en) | 2017-08-02 | 2017-08-02 | Vehicle type rough classification-based far-view scene vehicle detection method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107578048A true CN107578048A (en) | 2018-01-12 |
| CN107578048B CN107578048B (en) | 2020-11-13 |
Family
ID=61035431
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710653791.1A Active CN107578048B (en) | 2017-08-02 | 2017-08-02 | Vehicle type rough classification-based far-view scene vehicle detection method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107578048B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108364466A (en) * | 2018-02-11 | 2018-08-03 | 金陵科技学院 | A kind of statistical method of traffic flow based on unmanned plane traffic video |
| CN108470145A (en) * | 2018-01-31 | 2018-08-31 | 浙江工业大学 | A kind of vehicle steering wheel detection method based on slope of curve variation |
| CN109064495A (en) * | 2018-09-19 | 2018-12-21 | 东南大学 | A kind of bridge floor vehicle space time information acquisition methods based on Faster R-CNN and video technique |
| CN109272482A (en) * | 2018-07-20 | 2019-01-25 | 浙江浩腾电子科技股份有限公司 | A kind of urban road crossing vehicle queue detection system based on sequence image |
| CN109410598A (en) * | 2018-11-09 | 2019-03-01 | 浙江浩腾电子科技股份有限公司 | A kind of traffic intersection congestion detection method based on computer vision |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8503725B2 (en) * | 2010-04-15 | 2013-08-06 | National Chiao Tung University | Vehicle tracking system and tracking method thereof |
| CN103593981A (en) * | 2013-01-18 | 2014-02-19 | 西安通瑞新材料开发有限公司 | Vehicle model identification method based on video |
| CN104978567A (en) * | 2015-06-11 | 2015-10-14 | 武汉大千信息技术有限公司 | Vehicle detection method based on scenario classification |
| CN106529461A (en) * | 2016-11-07 | 2017-03-22 | 湖南源信光电科技有限公司 | Vehicle model identifying algorithm based on integral characteristic channel and SVM training device |
| CN106940784A (en) * | 2016-12-26 | 2017-07-11 | 无锡高新兴智能交通技术有限公司 | A kind of bus detection and recognition methods and system based on video |
-
2017
- 2017-08-02 CN CN201710653791.1A patent/CN107578048B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8503725B2 (en) * | 2010-04-15 | 2013-08-06 | National Chiao Tung University | Vehicle tracking system and tracking method thereof |
| CN103593981A (en) * | 2013-01-18 | 2014-02-19 | 西安通瑞新材料开发有限公司 | Vehicle model identification method based on video |
| CN104978567A (en) * | 2015-06-11 | 2015-10-14 | 武汉大千信息技术有限公司 | Vehicle detection method based on scenario classification |
| CN106529461A (en) * | 2016-11-07 | 2017-03-22 | 湖南源信光电科技有限公司 | Vehicle model identifying algorithm based on integral characteristic channel and SVM training device |
| CN106940784A (en) * | 2016-12-26 | 2017-07-11 | 无锡高新兴智能交通技术有限公司 | A kind of bus detection and recognition methods and system based on video |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108470145A (en) * | 2018-01-31 | 2018-08-31 | 浙江工业大学 | A kind of vehicle steering wheel detection method based on slope of curve variation |
| CN108470145B (en) * | 2018-01-31 | 2021-03-16 | 浙江工业大学 | Automobile steering wheel detection method based on curve slope change |
| CN108364466A (en) * | 2018-02-11 | 2018-08-03 | 金陵科技学院 | A kind of statistical method of traffic flow based on unmanned plane traffic video |
| CN108364466B (en) * | 2018-02-11 | 2021-01-26 | 金陵科技学院 | Traffic flow statistical method based on unmanned aerial vehicle traffic video |
| CN109272482A (en) * | 2018-07-20 | 2019-01-25 | 浙江浩腾电子科技股份有限公司 | A kind of urban road crossing vehicle queue detection system based on sequence image |
| CN109272482B (en) * | 2018-07-20 | 2021-08-24 | 浙江浩腾电子科技股份有限公司 | Urban intersection vehicle queuing detection system based on sequence images |
| CN109064495A (en) * | 2018-09-19 | 2018-12-21 | 东南大学 | A kind of bridge floor vehicle space time information acquisition methods based on Faster R-CNN and video technique |
| CN109064495B (en) * | 2018-09-19 | 2021-09-28 | 东南大学 | Bridge deck vehicle space-time information acquisition method based on fast R-CNN and video technology |
| CN109410598A (en) * | 2018-11-09 | 2019-03-01 | 浙江浩腾电子科技股份有限公司 | A kind of traffic intersection congestion detection method based on computer vision |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107578048B (en) | 2020-11-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wei et al. | Multi-vehicle detection algorithm through combining Harr and HOG features | |
| CN107578048A (en) | A kind of long sight scene vehicle checking method based on vehicle rough sort | |
| CN109977782B (en) | Cross-store operation behavior detection method based on target position information reasoning | |
| CN104933710B (en) | Based on the shop stream of people track intelligent analysis method under monitor video | |
| CN109190444B (en) | Method for realizing video-based toll lane vehicle feature recognition system | |
| CN101872416B (en) | Vehicle license plate recognition method and system of road image | |
| CN104951784A (en) | Method of detecting absence and coverage of license plate in real time | |
| CN105404857A (en) | Infrared-based night intelligent vehicle front pedestrian detection method | |
| EP2813973B1 (en) | Method and system for processing video image | |
| CN102819764A (en) | Method for counting pedestrian flow from multiple views under complex scene of traffic junction | |
| CN103400113B (en) | Freeway tunnel pedestrian detection method based on image procossing | |
| CN104978567A (en) | Vehicle detection method based on scenario classification | |
| CN113822285A (en) | Vehicle illegal parking identification method for complex application scene | |
| CN111008574A (en) | Key person track analysis method based on body shape recognition technology | |
| CN104463138A (en) | Text positioning method and system based on visual structure attribute | |
| CN112651293A (en) | Video detection method for road illegal stall setting event | |
| Chen et al. | A precise information extraction algorithm for lane lines | |
| CN110443142B (en) | Deep learning vehicle counting method based on road surface extraction and segmentation | |
| CN103927875A (en) | Traffic overflowing state recognition method based on video | |
| CN113095301B (en) | Road occupation operation monitoring method, system and server | |
| CN104331708B (en) | A kind of zebra crossing automatic detection analysis method and system | |
| CN110378935B (en) | Parabolic identification method based on image semantic information | |
| CN105069407B (en) | A kind of magnitude of traffic flow acquisition methods based on video | |
| CN106023270A (en) | Video vehicle detection method based on locally symmetric features | |
| Tan et al. | Shape template based side-view car detection algorithm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |